This study focuses on an automatic classification task aiming at placing community college students into the appropriate level (Level 1 and 2) of Developmental Education (DevEd) courses, according to their English L1 proficiency. DevEd courses are designed to remediate and support students' communication skills in reading and writing before they can fully participate in college-level or college-bearing courses. This paper uses machine-learning methods to investigate the impact of considering multiword expressions (MWE) as entire tokens on the automatic classification task. Since many MWE are often non-compositional in meaning and constitute a large percentage of the textual units of many texts, they are likely to have a relevant role in the data representation of texts and, hence, improve subsequent classification task. Information is scarce regarding the tokenization of MWE and how this affects automatic placement. To this end, a random, balanced corpus of 186 sample texts (93 from each level) was used. Experiments compared the performance of a set of classifiers on the plain text corpus and on a version of the same corpus annotated for MWE. Results showed that using MWE as lexical features improved the classification accuracy by 8.1% above the baseline.

The literature on second language learning posits that there are significant differences between the use of multiword expressions (MWE) by native speakers (NS) and non-native speakers (NNS). Furthermore, it considers that levels of language proficiency can be estimated on the basis of the use of these expressions. This paper analyses the written production from a corpus of essays written by native (16 essays, 5839 words) and non-native Spanish speakers (25 essays, 7767 words) enrolled in a course focused on the development of orthographic, grammatical, lexical, semantic, and discursive skills in Spanish. This is a required course for students pursuing a certification in Translating or Interpreting (Spanish/English) in the educational setting where the study took place. The corpus was manually tagged by two linguists. The classification scheme used was inspired by other schemes found in the literature and built for similar purposes. The results show that, in general, the distribution of MWE types found in the NS and NNS partition of the corpus was not very different (Pearson correlation: 0.894). However, interesting differences were found between the categories of verbal idioms and noun constructions. Though the corpus is too small for more significant conclusions to be drawn, it is possible